Curious AI #23
Oliver Rochford
Evangelist @ Auguria | Technologist | Cyberfuturist | Startup Advisor | Former Gartner Analyst
19 April 2024 - Welcome to the twenty-third issue of the Curious AI Newsletter, curated by cyber futurist and former Gartner Research Director Oliver Rochford, and synthesized by AI.?
Also check out the Intriguing Quantum Newsletter we just started by Daniella Pontes, CISSP , and me.?
Key Themes in this issue:
“There need to be broad breakthroughs, Any piece of efficiency matters.”
Rene Haas, ARM Holdings PLC CEO, on AI’s voracious appetite for energy (source )
Most Grounded
There’s an interesting article in the NY Times on how hard it is to verify the performance of AI systems like ChatGPT and Gemini without vigorous tests and testing standards. AI skills are outpacing established standards and tests that were designed for a past generation of models.?
Most curious
According to at least one tech executive, AI-powered virtual girlfriends might become a billion-dollar industry . It’s great that lonely people are finding solace through technology, but it’s also concerning, because the AI doesn’t possess emotions to truly reciprocate. It just reminds me too much of Jewel beetles mistaking a beer bottle for a mate. I also find it remarkable that it appears to be easier to convince someone an AI model is sentient, than to use AI to detect hackers.?
Weirdest
Someone had to announce an AI-driven beauty competition with digital women Aside from aising questions about body image and the ethics, what’s the point???
Must-read
15 Graphs That Explain the State of AI in 2024: The AI Index tracks the generative AI boom, model costs, and responsible AI use
IEEE Spectrum | https://spectrum.ieee.org/ai-index-2024
The 2024 AI Index Report reveals significant trends in AI development and adoption, showcasing the United States' dominance in AI research and the steep costs associated with training modern AI models. The report highlights the increasing global investment in AI, with the US leading, followed by China and the UK. It also emphasizes the escalating computational costs of AI, impacting the environment and resource allocation in the sector.
Sentiment: Neutral | Time to Impact: Immediate
Best humans still outperform artificial intelligence in a creative divergent thinking task
Humans are thought to be unique in their creativity. However, generative AI chatbots that can create high-quality artworks have emerged due to rapid AI progress, raising questions regarding human and machine creativity. In this study, we compared human creativity (n?=?256) against three AI chatbots using the widely used divergent thinking test, alternate uses task (AUT). Participants were invited to find unusual and inventive uses for everyday items. AI chatbots outperformed humans on average. Chatbots were more innovative than humans, who gave terrible suggestions. Best human ideas still outperformed chatbots. This study shows that AI can improve creativity, but it also emphasizes the unique and complicated aspect of human creativity, which AI may not be able to mimic or surpass. The study sheds light on human-machine creativity, which raises critical considerations concerning creative labor in the AI age.
Sentiment: Neutral | Time to Impact: Immediate
The Uneven Impact of Generative AI on Entrepreneurial Performance
Scalable and low-cost AI help is believed to boost company decision-making and economic performance.? Running a business involves many open-ended difficulties, making it difficult to extrapolate from recent studies showing that generative AI improves well-defined writing jobs. ? Our field experiment with 640 Kenyan entrepreneurs examined how AI-generated advise affected small business revenues and profits.A control group received a regular business guide, whereas a treatment group received a GPT-4-powered AI business mentor via WhatsApp. ? Our point estimates show that strong performers gained by slightly over 15% from AI coaching, while low performers did roughly 8% worse.? Exploratory WhatsApp interaction logs suggest that both groups sought the AI mentor's advise, but low performers scored worse since they sought aid on harder commercial tasks.? Our findings demonstrate generative AI's potential and limitations for global entrepreneurs.
Sentiment: Neutral | Time to Impact: Immediate
AI and Geopolitics
Europe is falling behind in generative AI, with the U.S. light-years ahead. But the race is just getting started
Fortune | https://fortune.com/europe/2024/04/15/europe-us-generative-ai-race-funding-brainstorm-ai-london/
Fortune reports on the competitive dynamics in the generative AI sector, highlighting the funding disparities between Europe and the US. At the Brainstorm AI conference in London, discussions focused on Europe's efforts to catch up through increased investment and regulatory frameworks. The article emphasizes the strategic importance of AI and the potential long-term benefits of Europe's unique approach to AI innovation and governance.
Sentiment: Neutral | Time to Impact: Mid-term
AI and Business
Baidu's AI Chatbot Ernie Bot Reaches 200 Million Users
Baidu's AI chatbot, Ernie Bot, has rapidly amassed 200 million users, demonstrating significant growth and adoption. This milestone highlights the expanding influence of AI technologies in everyday digital interactions and underscores Baidu's strong position in the competitive AI chatbot market. Ernie Bot's popularity reflects the increasing reliance on AI for communication and information services.
Sentiment: Positive | Time to Impact: Short-term
Humane AI Pin review roundup: an undercooked flop that's way ahead of its time
TechRadar | https://www.techradar.com/computing/artificial-intelligence/humane-ai-pin-review-roundup
Reviews of the Humane AI Pin, a wearable computer with an AI assistant, camera, and projector, paint it as an innovative but fundamentally flawed device. Despite its solid build and forward-thinking features, it disappoints with slow response times, overheating issues, and a high price tag. The consensus suggests it's not yet ready to replace smartphones or fulfill its potential in enhancing user interactions with AI.
Sentiment: Negative | Time to Impact: Mid-term
Humane's AI Pin got trashed by a YouTuber, and it sparked a firestorm over whether that's fair
Business Insider | https://www.businessinsider.com/youtuber-mkbh-dtrashes-humane-ai-pin-firestorm-over-damage-startup-2024-4
Tech reviewer Marques Brownlee, known as MKBHD, sparked controversy with his critical review of the Humane AI Pin, describing it as the "worst product" he's ever reviewed. This has intensified scrutiny on the startup behind the product, potentially impacting its reputation and future prospects. Brownlee's review highlights significant performance issues and a lack of practicality in the device.
Sentiment: Negative | Time to Impact: Short-term
AI and Efficient Government
领英推荐
State tax officials are using AI to go after wealthy payers
State tax officials are now employing AI tools to improve tax compliance among wealthy individuals. These advanced systems analyze vast amounts of financial data to identify inconsistencies and potential evasion, aiming to increase revenue and ensure fairness in tax collection. This use of AI represents a significant shift towards more data-driven governance in state tax administration.
Sentiment: Positive | Time to Impact: Short-term
AI Carbon Footprint
AI Computing Is on Pace to Consume More Energy Than India, Arm Says
Arm Holdings predicts that by 2030, AI computing will consume more electricity than India, highlighting the urgent need for energy-efficient technology solutions. As AI's capabilities expand, the industry must innovate to manage the soaring energy demands of data centers, potentially through the adoption of custom-built, energy-efficient Arm chips.
Sentiment: Concerned | Time to Impact: Mid-term
AI could gobble up a quarter of all electricity in the U.S. by 2030 if it doesn’t break its energy addiction
As AI technologies like ChatGPT advance, they're driving up the demand for more powerful and efficient semiconductors. Companies are increasingly turning to ARM chips for their energy efficiency, crucial as AI applications, particularly large language models, escalate global electricity usage. The shift underscores the urgent need for technological innovations that reduce energy consumption without stifling AI's potential.
Sentiment: Neutral | Time to Impact: Short-term
AI and Society
Tech exec predicts ‘AI girlfriends’ will create $1B business: ‘Comfort at the end of the day’
New York Post | https://nypost.com/2024/04/15/business/tech-exec-predicts-ai-girlfriends-will-create-1b-business/
A tech executive predicts that AI-powered virtual girlfriends could become a billion-dollar industry. These digital companions, designed to offer conversation and emotional engagement, leverage advanced AI algorithms to simulate realistic interactions. This development could revolutionize social dynamics and open up new avenues for AI application in personal relationships.
Sentiment: Positive | Time to Impact: Mid-term
World's first beauty pageant for AI women is announced: 'Miss AI' contest will see computer-generated ladies face off in tests of beauty
Daily Mail | https://www.dailymail.co.uk/sciencetech/article-13302163/beauty-pageant-AI-women-announced.html
An AI-driven beauty pageant has been announced, featuring digital representations of women, marking a controversial step in the blending of technology with traditional beauty standards. This virtual contest, utilizing sophisticated AI models to create and judge entrants, has sparked discussions on the implications for body image and the ethical aspects of digital personifications in competitions.
Sentiment: Neutral | Time to Impact: Mid-term
AI in Academia
Is ChatGPT corrupting peer review? Telltale words hint at AI use
Researchers are increasingly using AI tools like ChatGPT for peer review, potentially affecting the quality of feedback in scientific publishing. A study indicates that up to 17% of peer reviews may be influenced by AI, raising concerns about accuracy and integrity, given AI's tendency to "hallucinate." The discussion highlights the need for transparency in AI's role in peer reviews to maintain accountability in academic evaluations.
Sentiment: Neutral | Time to Impact: Immediate
AI and the path to AGI
AI just beat a human test for creativity. What does that even mean?
MIT Technology Review | https://www.technologyreview.com/2023/09/14/1079465/ai-just-beat-a-human-test-for-creativity-what-does-that-even-mean/
MIT Technology Review discusses an AI's achievement in outperforming humans on a creativity test, raising questions about the definition and assessment of creativity in AI. The article explores the implications of this milestone and the challenges in measuring AI creativity against human standards, emphasizing the evolving nature of AI capabilities and the potential for redefining creative benchmarks in technology.
Sentiment: Neutral | Time to Impact: Long-term
A.I. Has a Measurement Problem
The New York Times | https://www.nytimes.com/2024/04/15/technology/ai-models-measurement.html
The lack of standardized testing for AI systems like ChatGPT and Gemini makes it difficult to verify their capabilities, posing risks to users who rely on AI company claims alone. The AI Index Report by Stanford highlights this as a major challenge, emphasizing the need for robust, systematic evaluations to compare AI models accurately. Innovative tests are required as AI capabilities rapidly evolve, surpassing traditional benchmarks.
Sentiment: Neutral | Time to Impact: Mid-term
AI and Cybersecurity
OpenAI's GPT-4 can exploit real vulnerabilities by reading security advisories
Researchers have demonstrated that GPT-4 can identify and exploit real-world cybersecurity vulnerabilities. This revelation underscores the potential risks associated with advanced AI models, prompting discussions on the need for ethical guidelines and enhanced security measures to prevent misuse of AI capabilities in cybersecurity.
Sentiment: Negative | Time to Impact: Short-term
About the Curious AI Newsletter
AI is hype. AI is a utopia. AI is a dystopia.
These are the narratives currently being told about AI. There are mixed signals for each scenario. The truth will lie somewhere in between. This newsletter provides a curated overview of positive and negative data points to support decision-makers in forecasts and horizon scanning. The selection of news items is intended to provide a cross-section of articles from across the spectrum of AI optimists, AI realists, and AI pessimists and showcase the impact of AI across different domains and fields.
The news is curated by Oliver Rochford , Technologist, and former Gartner Research Director. AI (ChatGPT) is used in analysis and for summaries.
Want to summarize your news articles using ChatGPT? Here's the latest iteration of the prompt. The Curious AI Newsletter is brought to you by the Cyber Futurists .